Description

Amid the growing accessibility of performant Large Language Models (LLMs) like GPT-4 and Llama-2, and a burgeoning range of commercial licenses, enterprises are now forging the first wave of LLM-driven applications. This session will begin by addressing the challenges with LLMs such as defining clear success criteria for an LLM project and understanding different approaches to LLM work. 

We’ll further address technical challenges such as memory handling, input quality control ("garbage in, garbage out"), and the complexities of embeddings. A comparison of different approaches from using Retrieval-Augmented Generation (RAG) to training or fine-tuning on top of a variety of different LLMs. 

Through real-world examples we will discuss practical approaches to risk management, ethical considerations, and the ideal team composition for an LLM project. In addition, we’ll discuss a variety of tools that form the modern LLM application development stack. Join us to demystify LLM applications and equip your organization with the knowledge to succeed.


Local ODSC chapter in NYC, USA

Instructor's Bio

Nicolas Decavel-Bueff

Data Science Consultant at Pandata

He is an SF-based Data Scientist who has delivered valuable solutions across a broad spectrum of industries. His accomplishments range from employing natural language processing models in logistics to leading teams in the development of vital models in the utility sector. His work with diverse tools has consistently created quantifiable business impact. Equipped with a Master's in Data Science from the University of San Francisco, Nicolas blends academic rigor and practical experience to address complex business challenges.

Parham Parvizi

Founder of Data Stack Academy and Tura.io

Parham is a founding member of Tura.io and DataStack.Academy. Tura is a group of professional Cloud Data Engineers and Architects while Data Stack Academy is the most comprehensive Data Engineering bootcamp; training the future of Cloud Data Engineers. In his 20 years as Data Engineer and Cloud/Big Data Solution Architect, he has been an Apache Software Foundation contributor and an early adopter and contributor to open source Big Data projects as Map Reduce and Hive.  Prior to Tura Labs, he was a product manager at Pivotal and one of the initial members of Talend. Prior to Tura Labs, he was a product manager at Pivotal and one of the initial members of Talend. As a Data Advisor and consultant, Parham’s has had the opportunity  and pleasure to work with nearly every fortune 100 company over the years. From managing thousands node clusters to optimizing data task that you are familiar with behind the scenes.

Cal Al-Dhubaib

Founder & AI Strategist at Pandata

Cal is a globally recognized data scientist, entrepreneur, and innovator in responsible artificial intelligence, specializing in high-risk sectors such as healthcare, energy, and defense. He is the founder and CEO of Pandata, a consulting company that helps organizations to design and develop AI-driven solutions for complex business challenges, with an emphasis on responsible AI. Their clients include globally recognized organizations like the Cleveland Clinic, Progressive Insurance, University Hospitals, and Parker Hannifin. Cal frequently speaks on topics including AI ethics, change management, data literacy, and the unique challenges of implementing AI solutions in high-risk industries. His insights have been featured in noteworthy publications such as Forbes, Ohiox, the Marketing AI Institute, Open Data Science, and AI Business News. Cal has also received recognition among Crain’s Cleveland Notable Immigrant Leaders, Notable Entrepreneurs, and most recently, Notable Technology Executives.

Webinar

  • 1

    ON-DEMAND WEBINAR: "Preparing for your First Enterprise Large Language Model (LLM) Application"

    • Ai+ Training

    • Webinar recording

UPCOMING LIVE TRAINING

Register now to save 30%